Here's the deal with prototyping and customizing shoes for elites. They start wearing different versions of the shoe and then give feedback as to what they do and don't like. That feedback is taken and built into the next prototype. This can go on until they feel like they have a version customized and 'perfected' for a specific athlete. One they have a 'perfected' sample they may build a brand new pair based on those same specifications for race day. If you notice now all of the Streak 6 models have glued in insoles. The ones that Kipchoge wore may have been brand new and the spray glue that was used may have not gotten a bond and the insoles then worked free. He then went on to wear Vaporfly 4% prototypes in London and Rio with no issues.
Take Galen, for example. The shoes that he wore at Boston were modified for him between Prague and Boston.
Bekele had certainly tested shoes that were exact copies of the ones that he wore at London yesterday, but they made him a brand new pair for race day. Maybe he did something like a 10K tempo effort in the prototype and deemed them suitable, but maybe you aren't going to have a blister until you hit 15K at race pace instead. It could be something very simple. A rough seam, the race course and terrain, or temperatures and conditions on race day. Do you have any idea how many tiny variables can affect performance in a shoe? It's very hard to get it 100% right when you are customizing shoes for an athlete.
At any rate it's unfortunate that his shoes detracted from his performance. He ran a ballsy race and I thought that he might still have a shot at the win with half a mile to go. Hopefully he recovers well and we will possibly see him in the World Championships.
The Vaporfly Elite has had great success so far and this was an incident where it just wasn't tuned quite right for the specific athlete. The wear testing team works really hard to take the feedback from each elite athlete and to get the shoe 100% right for them, but it's not always a perfect process.